Synthesis and biochemical characterization of quasi-stable trimer models of full-length amyloid β40 with a toxic conformation

Article abstract of DOI:10.1039/c8cc08618d

Here, we report the first synthesis of quasi-stable trimer models of full-length Aβ40 with a toxic conformation using a 1,3,5-phenyltris-l-alanyl linker at position 34, 36, or 38. The only trimer to exhibit weak neurotoxicity against SH-SY5Y cells was the one which was linked at position 38. This suggests that such a propeller-type trimer model is not prone to forming oligomers with potent neurotoxicity, which is in contrast with its corresponding dimer model.

Full text of DOI:10.1039/c8cc08618d

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Synthesis and biochemical characterization of
quasi-stable trimer models of full-length amyloid
b40 with a toxic conformation†
a
Cite this: Chem. Commun., 2019,
55, 182
Received 28th October 2018,
Accepted 30th November 2018
Yumi Irie,a Mizuho Hanaki,a Kazuma Murakami,
Tsuneo Imamoto,b
Takumi Furuta,c Takeo Kawabata,
Ken-ichi Akagi,e Ryo C. Yanagita
Taiji Kawase,d Kenji Hirose,d Yoko Monobe,e
c
f
a
DOI: 10.1039/c8cc08618d
and Kazuhiro Irie
*
rsc.li/chemcomm
Here, we report the first synthesis of quasi-stable trimer models of full- states of Ab, photochemically cross-linked Ab oligomers have also
length Ab40 with a toxic conformation using a 1,3,5-phenyltris-L-alanyl been prepared and were found to be neurotoxic, but the cross-
linker at position 34, 36, or 38. The only trimer to exhibit weak linking positions were not limited to one residue.14–17
neurotoxicity against SH-SY5Y cells was the one which was linked at
We have recently synthesized quasi-stable dimers of E22P–
position 38. This suggests that such a propeller-type trimer model is Ab40 and E22P–Ab42 with an L,L-2,6-diaminopimeric acid (DAP)
not prone to forming oligomers with potent neurotoxicity, which linker at positions 38 and 40, respectively, and in doing so, have
is in contrast with its corresponding dimer model.
demonstrated that the formation of the C-terminal hydro-
phobic core is necessary to form toxic oligomers.9,18 Since the
Amyloid b proteins are peptides of 36–43 amino acids in length E22P mutation could accelerate the formation of the toxic
(Ab36–Ab43) found in senile plaques1,2 that are involved in oligomers,19–21 we named the turn structure at positions 22
the pathogenesis of Alzheimer’s disease (AD).3,4 Oligomeric and 23 the ‘‘toxic turn’’. These two dimer models existed as
assemblies of Ab could induce synaptotoxicity during AD quasi-stable oligomers, and exhibited potent neurotoxicity
progression.5 Ab42 is the protein which is most prone to against SH-SY5Y cells.9,18 Although oligomers derived from
aggregation and to forming toxic oligomers as well as less-toxic Ab dimers could be involved in the pathogenesis of AD,22–24
fibrils, but the oligomers of Ab40 can also become neurotoxic as several reports have suggested that Ab trimers as well as Ab
demonstrated by the dimer models of Ab40.6–9 ‘‘Aggregation’’ dimers could also contribute to its onset.25,26 Recently, the
in this paper is defined as the change of Ab monomers into synthesis of trimer models of Ab fragments has been reported
oligomers, protofibrils, or amyloid fibrils. It is quite difficult to and characterized,27,28 however, there are no reports describing
determine the mechanism of toxicity of Ab oligomers since the the synthesis of full-length Ab trimer models. In this paper, we
aggregation speed of Ab monomers is very fast. The oligomeric report the first synthesis and biochemical characterization of
state could not easily be captured even when using ion mobility- quasi-stable trimer models of full-length Ab40 with a toxic
mass spectrometry (IM-MS).10,11 To overcome this difficulty, conformation using a 1,3,5-phenyltris-L-alanyl linker (Fig. 1A)
various dimer models of Ab have been synthesized and at position 34, 36, or 38 (7–9 in Fig. 1C). Our goal in doing this
characterized.6–9,12,13 These studies identified the important resi- was to use these models to investigate whether the dimers or
dues involved in the oligomerization and fibrillization processes. trimers are more relevant to AD progression by comparing the
As shown by Teplow’s pioneering studies to capture oligomeric E22P–Ab40 dimer model with a DAP linker at position 38 (10).
We selected E22P–Ab40 as a template for the full-length trimer
because the aggregative ability and neurotoxicity of wild-type
a Division of Food Science and Biotechnology, Graduate School of Agriculture,
Ab40 are much lower than those of E22P–Ab40.29
Kyoto University, Kyoto 606-8502, Japan. E-mail: irie@kais.kyoto-u.ac.jp;
We adopted a 1,3,5-phenyltris-L-alanyl linker for the trimer
Fax: +81-75-753-6284; Tel: +81-75-753-6281
synthesis since this linker enables the formation of anti-parallel
b-sheets at the C-terminal region, which exists in one of the
proposed trimer models by Huang et al. using solid-state NMR
[Fig. 1B(i, ii)].30 A molecular modelling study indicated that the
V36 core region of this trimer model is reproducible by linking
position 36 with the 1,3,5-phenyltris-L-alanyl linker [Fig. 1B(iii)].
We therefore synthesized (S,S,S)-Fmoc-1,3,5-phenyltris-L-alanine
to be used in solid-phase synthesis as one amino acid residue
b Nippon Chemical Industrial Co. Ltd, Tokyo 136-8515, Japan
c Institute for Chemical Research, Kyoto University, Kyoto 611-0011, Japan
d Nihon Waters K. K., Tokyo 140-0001, Japan
e National Institute of Biomedical Innovation, Health and Nutrition,
Osaka 567-0085, Japan
f Department of Applied Biological Science, Faculty of Agriculture,
Kagawa University, Kagawa 761-0795, Japan
† Electronic supplementary information (ESI) available: Experimental procedures
and Fig. S1–S3. See DOI: 10.1039/c8cc08618d
182 | Chem. Commun., 2019, 55, 182--185
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Communication
ChemComm
with modifications (Fig. 1A). Horner–Wadsworth–Emmons olefi-
nation of 1 gave 3, whose stereochemistry (all-Z) was confirmed by
the NOESY spectrum.
Asymmetric hydrogenation of 3 using Rh-(S,S)-QuinoxP32
gave 4 with a 92% yield. The diastereomeric ratio and enantio-
meric excess of 4, after purification using a silica-gel column,
were determined to be 498% (Fig. S1 in ESI†). Alkaline hydro-
lysis of 4, followed by deprotection, gave 5, whose amino groups
were protected with Fmoc. The overall yield of the linker 6
was 25%.
The three trimer models of E22P–Ab40 with the linker 6 at
position 34, 36, or 38 (7–9, Fig. 1C) were synthesized in a
stepwise fashion on preloaded Fmoc-L-Val-PEG-PS resin by a
microwave peptide synthesizer (ESI†). Each coupling reaction
was carried out using each Fmoc amino acid, HATU,33 and
N,N-diisopropylethylamine in DMF. The 1/3 molar equivalent
trimer linker 6 (0.033 mmol) was employed instead of Fmoc-L-Leu,
Fmoc-L-Val, or Fmoc-Gly at position 34, 36, or 38, respectively, in
order to avoid formation of the mono-coupled and/or di-coupled
peptides. This procedure gave 7–9 at yields of 0.58%, 1.1%, and
0.45%, respectively, after HPLC purification using C4 and C18
columns under acidic conditions (ESI†). Their purity was checked
by HPLC analysis, and their molecular weights and formulae were
confirmed by ESI-qTOF-MS measurements (Fig. S2 in ESI†). Only
the data of 8 are shown in Fig. 1D as a representative example.
The aggregative ability of the trimer models (7–9) was evalu-
ated using thioflavin-T (Th-T) that fluoresces when bound to the
b-sheet structure in Ab aggregates (Fig. 2A).34 Unlike wild-type
Ab40,29 E22P–Ab40 with the toxic turn promptly aggregated to give
a high level of fluorescence. In contrast, the fluorescence intensity
of 7–9 was weaker than that of E22P–Ab40 for two weeks,
suggesting that these trimer models did not form typical fibrils like
E22P–Ab40. This was confirmed by transmission electron micro-
scopy (TEM) analysis (Fig. 2B). Only small amounts of amorphous
aggregates were detected for 7 and 8, while 9 and 10 formed
protofibrils (fibrils) after 48 h. E22P–Ab40 showed the highest
fluorescence intensity among them. The fluorescence intensity in
the Th-T assay was different among E22P–Ab40, 9, and 10; this
might reflect the number of fibrils and their structural differences.
Furthermore, time-dependent circular dichroism (CD)
measurements were conducted (Fig. 2C). E22P–Ab40 exhibited
a positive peak at ca. 195 nm and a negative peak at ca. 215 nm
after 4 h incubation, indicating the formation of the b-sheet
structure. In contrast, the secondary structures of 7 and 8 remained
almost random even after 48 h incubation. The trimer 9 with the
linker 6 at position 38 formed a b-sheet rich structure in a time-
dependent manner. Consistently, the TEM analyses after 48 h
incubation (Fig. 2B) showed that only 9 formed protofibrils (fibrils).
We then measured the neurotoxicity of 7–9 against SH-SY5Y
cells (human neuroblastoma) using the MTT [(4,5-dimethylthiazol-
2-yl)-2,5-diphenyltetrazolium bromide] assay (Fig. 3A). After incuba-
tion for 16 h with each compound at various concentrations,
cell viability was estimated by the ability to reduce MTT. The dimer
model 10 significantly decreased cell viability even at 1 mM. Among
Fig. 1 (A) Synthesis of (S,S,S)-Fmoc-1,3,5-phenyltris-L-alanine: (a) 1,1,3,3-
tetramethyl guanidine, THF, 0 1C to rt, 4 h, 76%. (b) Rh-(S,S)-QuinoxP,32
H2 (4 atm), EtOAc/MeOH, rt, 4 h, 92% (498%ee, 498%de). (c) NaOHaq, MeOH,
1 h, 97%. (d) H2, Pd/C, MeOH, overnight, 75%. (e) Fmoc-OSu, Na2CO3, MeCN,
H2O, 3 d, 50%. (B) (i) Idealized trimer model in the C-terminal region (K28–A42)
of wild-type Ab42 based on solid-phase NMR proposed by Huang et al.30
(ii) Snapshot of the Ab42 trimer model in the C-terminal region after 1 ns of
molecular dynamics simulation. (iii) The C-terminal region of the trimer model 8
with the 1,3,5-phenyltris-L-alanyl linker at position 36; the snapshot was
extracted after 1 ns of molecular dynamics simulation. In A–C, ribbons represent
anti-parallel b-strands, and V36 and the phenyl linker residues are represented in
stick models and coloured orange. For clarity, nonpolar hydrogen atoms are not
displayed. (C) Structure of the trimer models synthesized in this study (7–9)
along with the dimer model (10).9 (D) HPLC profile and ESI-qTOF-MS data with
deconvolution of the trimer model 8 as a representative example. HPLC
condition: X-bridge (100 Â 4.6 mm I.D.), 1 mL minÀ1, UV 220 nm, 20–50%
acetonitrile containing 0.1% trifluoroacetic acid (30 min linear gradient), 10 mg/
10 mL (0.15% NH4OH) injection. 8, m/z 12880.62 (calcd for av. mass, 12881.58).
at position 34, 36, or 38. The synthesis started with 1,3,5-tri- the trimer models (7–9), only 9 exhibited weak neurotoxicity
formylbenzene (1) according to previously reported methods,31 through somewhat decreased cell viability.
This journal is ©The Royal Society of Chemistry 2019
Chem. Commun., 2019, 55, 182--185 | 183

View Article Online
ChemComm
Communication
Fig. 3 (A) Neurotoxicity against SH-SY5Y of the trimer models 7–9 along
with that of E22P–Ab40 monomer and the dimer model 10 after 16 h
incubation at 37 1C. Absorbance obtained after adding vehicle (0.15%
NH4OH) was taken as 100%. *p o 0.05, #p o 0.05 versus vehicle.
(B) IM-MS of 7–9 (8 mM) along with 10 (12.5 mM) after 4 h incubation at
37 1C. ‘‘n’’ denotes an integer corresponding to the number of units
coexisting in the solution depending on their drift time.
Fig. 2 (A) Th-T fluorescence of the trimer models 7–9 (8 mM) along with
that of E22P–Ab40 (25 mM) and the dimer model 10 (12.5 mM) after
the indicated incubation times at 37 1C. Data are expressed as mean Æ
SD (n = 8). (B) TEM analyses of the aggregates of 7–10, and E22P–Ab40
(8, 12.5, or 25 mM) after 48 h incubation at 37 1C. Scale bar = 50 nm
(magnification: 30 k). (C) CD spectra of 7–9 (8 mM) and E22P–Ab40 (25 mM)
after incubation for the indicated duration at 37 1C.
are due to differences in oligomer size, or due to other
consequences of the different chemistries of the models, for
example, the extent of oligomer/fibril formation, oligomer/fibril
Finally, in order to determine the size of the oligomers structure, or distinct oligomer/fibril surface chemistry. These
ascribable to the neurotoxicity, measurements were conducted issues may be clarified by further studies on the structural basis
using IM–MS combined with nanoelectrospray. Since IM–MS of these models using NMR and cryo-EM.
after 2, 4, and 6 h incubation was similar in each trimer model,
To the best of our knowledge, this is the first report on the
only the 2D heat maps with m/z and drift time after 4 h synthesis and biochemical characterization of the full-length
incubation were shown in Fig. 3B. Their projection of the 2D trimer model based on the trimer structure in Ab42 oligomers
spectra on the mass spectra axis is shown in Fig. S3 in ESI.† (150 kDa, ca. 30-mer) deduced from solid-state NMR.30 This
After deconvolution based on the observed and calculated trimer model with the linker at position 36 (8) was quasi-stable,
masses, cross-peaks were assigned to the series of multivalent but its neurotoxicity was very weak compared with the dimer
ions (n = 1, 2, 3. . . of each trimer model denotes trimer, and the trimer models with the linkers at position 38 (10 and 9).
hexamer, nonamer, . . ., respectively). As a control, the dimer These data suggest that the position of the C-terminal hydro-
model 10 with a DAP linker at position 38 formed higher-order phobic core could influence the toxicity of Ab oligomers, and
oligomers (n = 6–12 : 12–24-mer), and the E22P–Ab40 monomer that at least the propeller-like trimer model 9 is not prone to
did not give any cross peaks because of high aggregation form oligomers with potent neurotoxicity, which is in contrast
velocity (data not shown). Trimer 7 with the linker at position with its corresponding dimer model 10. Although the trimer
34 did not form oligomers; it remained a trimer and hexamer models in this study reference the Ab40 protein, the C-terminal
(n = 1–2) even after 24 h of incubation (data not shown). hydrophobic interaction in the Ab42 dimer could partly be
In contrast, 8 existed as a 3–12-mer (n = 1–4), and 9 formed mimicked by the Ab40 dimer model via the covalent linkage
larger oligomers of 9–21 mer (n = 3–7). It should be stated that it at the C-terminal region; in fact, the neurotoxicities of
remains unclear if the distinct toxicities of the Ab40 trimers the dimer models of Ab40 and Ab42 with the DAP linker at
184 | Chem. Commun., 2019, 55, 182--185
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Communication
ChemComm
7 W. M. Kok, J. M. Cottam, G. D. Ciccotosto, L. A. Miles, J. A. Karas,
D. B. Scanlon, B. R. Roberts, M. W. Parker, R. Cappai, K. J. Barnham
and C. A. Hutton, Chem. Sci., 2013, 4, 4449–4454.
8 T. T. O’Malley, N. A. Oktaviani, D. Zhang, A. Lomakin, B. O’Nuallain,
S. Linse, G. B. Benedek, M. J. Rowan, F. A. A. Mulder and
D. M. Walsh, Biochem. J., 2014, 461, 413–426.
9 Y. Irie, K. Murakami, M. Hanaki, Y. Hanaki, T. Suzuki, Y. Monobe,
T. Takai, K.-i. Akagi, T. Kawase, K. Hirose and K. Irie, ACS Chem.
Neurosci., 2017, 8, 807–816.
10 S. L. Bernstein, N. F. Dupuis, N. D. Lazo, T. Wyttenbach, M. M.
Condron, G. Bitan, D. B. Teplow, J. E. Shea, B. T. Ruotolo,
C. V. Robinson and M. T. Bowers, Nat. Chem., 2009, 1, 326–331.
11 M. Kloniecki, A. Jablonowska, J. Poznanski, J. Langridge, C. Hughes,
I. Campuzano, K. Giles and M. Dadlez, J. Mol. Biol., 2011, 407,
110–124.
12 W. M. Kok, D. B. Scanlon, J. A. Karas, L. A. Miles, D. J. Tew,
M. W. Parker, K. J. Barnham and C. A. Hutton, Chem. Commun.,
2009, 6228–6230.
13 T. Yamaguchi, H. Yagi, Y. Goto, K. Matsuzaki and M. Hoshino,
Biochemistry, 2010, 49, 7100–7107.
14 G. Bitan, A. Lomakin and D. B. Teplow, J. Biol. Chem., 2001, 276,
35176–35184.
15 G. Bitan and D. B. Teplow, Acc. Chem. Res., 2004, 37, 357–364.
16 D. B. Teplow, N. D. Lazo, G. Bitan, S. Bernstein, T. Wyttenbach,
M. T. Bowers, A. Baumketner, J.-E. Shea, B. Urbanc, L. Cruz,
J. Borreguero and H. E. Stanley, Acc. Chem. Res., 2006, 39, 635–645.
17 G. Yamin, T.-P. V. Huynh and D. B. Teplow, Biochemistry, 2015, 54,
5315–5321.
18 K. Murakami, M. Tokuda, T. Suzuki, Y. Irie, M. Hanaki, N. Izuo,
Y. Monobe, K.-i. Akagi, R. Ishii, H. Tatebe, T. Tokuda, M. Maeda,
T. Kume, T. Shimizu and K. Irie, Sci. Rep., 2016, 6, 29038.
19 A. Morimoto, K. Irie, K. Murakami, Y. Masuda, H. Ohigashi,
M. Nagao, H. Fukuda, T. Shimizu and T. Shirasawa, J. Biol. Chem.,
2004, 279, 52781–52788.
20 K. Murakami, K. Irie, H. Ohigashi, H. Hara, M. Nagao, T. Shimizu
and T. Shirasawa, J. Am. Chem. Soc., 2005, 127, 15168–15174.
21 Y. Masuda, S. Uemura, R. Ohashi, A. Nakanishi, K. Takegoshi,
T. Shimizu, T. Shirasawa and K. Irie, ChemBioChem, 2009, 10,
287–295.
positions 38 and 40, respectively, were quite similar to each
other.9,18
In summary, we have synthesized three full-length trimer
models, 7–9, with toxic conformations. The linker 6 was selected to
reproduce the putative trimer structure of Ab42 oligomers.30 The
synthesis of these models is straightforward and practical although
the purification step involves some laborious operations. Only 9,
with the linker at position 38, formed larger oligomers of 9–21-mer
that were weakly neurotoxic against SH-SY5Y cell lines. Therefore,
the trimer model 9 as well as the dimer model 10 may be useful
tools to understand amyloid aggregation and how the aggregates
correlates to toxicity.
Notably, as the neurotoxicities of 7–9 did not exceed that of
the monomer and the dimer model 10, which has an inter-
molecular parallel b-sheet, these propeller-type trimer models
with an anti-parallel b-sheet at the C-terminal region by the
linker 6 may not reflect the structure of toxic Ab oligomers.
Alternatively, Ab dimers like 10 might play a more critical role
in the pathogenesis of AD compared with the propeller-type
trimers. In order to verify which is more important, further
attempts to synthesize other types of trimer models, for example
trimers with intermolecular parallel b-sheets similar to the dimer
model 10, are required. Synthesis of the corresponding 42-mer
models would also be necessary.
It is quite difficult to determine the mechanism of toxicity of
Ab oligomers since the aggregation speed of Ab monomers is
very fast. In particular, while this is true for in vivo conditions,
studies have shown the mechanisms of toxicity via membrane
disruption and two steps of pore formation versus fiber-
dependent detergent-like mechanisms.35 Since synthetic oligo-
mers do not always reflect the heterogeneous mixture present
in vivo, the dynamics (or mobility) of residues that may be
responsible for toxicity could be different for synthetic species.
In addition, their affinity for the cell membrane and its
disruption could also be different.
22 C. Lendel, M. Bjerring, A. Dubnovitsky, R. T. Kelly, A. Filippov,
¨
O. N. Antzutkin, N. C. Nielsen and T. Hard, Angew. Chem., Int. Ed.,
2014, 53, 12756–12760.
23 I. Klyubin, V. Betts, A. T. Wetzel, K. Blennow, H. Zetterberg,
A. Wallin, C. A. Lemere, W. K. Cullen, Y. Peng, T. Wisniewski, D. J.
Selkoe, R. Anwyl, D. M. Walsh and M. J. Rowan, J. Neurosci., 2008,
28, 4231–4237.
24 G. M. Shankar, S. Li, T. H. Mehta, A. Garcia-Munoz, N. E.
Shepardson, I. Smith, F. M. Brett, M. A. Farrell, M. J. Rowan,
C. A. Lemere, C. M. Regan, D. M. Walsh, B. L. Sabatini and
D. J. Selkoe, Nat. Med., 2008, 14, 837–842.
25 T. Tomiyama, T. Nagata, H. Shimada, R. Teraoka, A. Fukushima,
H. Kanemitsu, H. Takuma, R. Kuwano, M. Imagawa, S. Ataka,
Y. Wada, E. Yoshioka, T. Nishizaki, Y. Watanabe and H. Mori,
Ann. Neurol., 2008, 63, 377–387.
This work was supported by JSPS KAKENHI Grant Number
26221202 to K. I. and K. M. The authors thank Ms Tomoyo
Takai and Ms Ritsuko Yamaguchi at the National Institute of
Biomedical Innovation for their assistance with TEM analysis.
26 M. Townsend, G. M. Shankar, T. Mehta, D. M. Walsh and D. J. Selkoe,
J. Physiol., 2006, 572, 477–492.
27 R. K. Spencer, H. Li and J. S. Nowick, J. Am. Chem. Soc., 2014, 136,
5595–5598.
28 K. Shinoda, Y. Sohma and M. Kanai, Bioorg. Med. Chem. Lett., 2015,
25, 2976–2979.
Conflicts of interest
There are no conflicts to declare.
29 K. Murakami, K. Irie, A. Morimoto, H. Ohigashi, M. Shindo,
M. Nagao, T. Shimizu and T. Shirasawa, J. Biol. Chem., 2003, 278,
46179–46187.
30 D. Huang, M. I. Zimmerman, P. K. Martin, A. J. Nix, T. L. Rosenberry
and A. K. Paravastu, J. Mol. Biol., 2015, 427, 2319–2328.
Notes and references
1 G. G. Glenner and C. W. Wong, Biochem. Biophys. Res. Commun.,
1984, 120, 885–890.
2 C. L. Masters, G. Simms, N. A. Weinman, G. Multhaup, B. L.
´
McDonald and K. Beyreuther, Proc. Natl. Acad. Sci. U. S. A., 1985, 31 A. Ritzen, B. Basu, A. Wållberg and T. Frejd, Tetrahedron: Asymmetry,
82, 4245–4249.
1998, 3491–3496.
3 J. A. Hardy and G. A. Higgins, Science, 1992, 256, 184–185.
4 J. Hardy and D. J. Selkoe, Science, 2002, 297, 353–356.
5 D. M. Walsh, I. Klyubin, J. V. Fadeeva, W. K. Cullen, R. Anwyl, M. S.
Wolfe, M. J. Rowan and D. J. Selkoe, Nature, 2002, 416, 535–539.
32 T. Imamoto, K. Tamura, Z. Zhang, Y. Horiuchi, M. Sugiya,
K. Yoshida, A. Yanagisawa and I. D. Gridnev, J. Am. Chem. Soc.,
2012, 134, 1754–1769.
33 L. A. Carpino, J. Am. Chem. Soc., 1993, 115, 4397–4398.
6 B. O’Nuallain, D. B. Freir, A. J. Nicoll, E. Risse, N. Ferguson, 34 H. Naiki and F. Gejyo, Methods Enzymol., 1999, 309, 305–318.
C. E. Herron, J. Collinge and D. M. Walsh, J. Neurosci., 2010, 30, 35 S. A. Kotler, P. Walsh, J. R. Brender and A. Ramamoorthy, Chem. Soc.
14411–14419.
Rev., 2014, 43, 6692–6700.
This journal is ©The Royal Society of Chemistry 2019
Chem. Commun., 2019, 55, 182--185 | 185